Legal claims defining the scope of protection, as filed with the USPTO.
4. The method of claim 3, wherein said one or more avatars comprise at least first and second avatars, and wherein said one or more negotiation software agents comprise at least corresponding first and second negotiation software agents, further comprising determining that said human speech signals and said non-speech behavior are directed to said at least one of said one or more avatars.
5. The method of claim 4, further comprising, responsive to said approving, broadcasting details of said proposed negotiation act to said at least first and second negotiation software agents.
7. The method of claim 6, wherein identifying said closest one of said one or more avatars to said apparent attention spot as said at least one of said one or more avatars to which said human speech is directed is based on a distance between said closest one of said one or more avatars and said apparent attention spot being less than a threshold value.
8. The method of claim 6, wherein said apparent attention spot varies during an utterance, and wherein identifying said closest one of said one or more avatars to said apparent attention spot as said at least one of said one or more avatars to which said human speech is directed is based on said at least one of said one or more avatars to which said human speech is directed being closest to said attention spot during a greatest fraction of said utterance.
9. The method of claim 6, further comprising highlighting said closest one of said one or more avatars to said apparent attention spot.
10. The method of claim 5, wherein determining that said human speech signals and said non-speech behavior are directed to said at least one of said one or more avatars comprises using a trained addressee classification model to classify a head orientation time series of a head of a human negotiator who utters said human speech signals and engages in said non-speech behavior into a time series of inferred attention avatars that identifies said at least one of said one or more avatars to which said human speech signals and said non-speech behavior are directed.
11. The method of claim 10, wherein said classification model employs deep learning.
12. The method of claim 5, wherein said approving of said first proposed negotiation act and said disapproving of said second proposed negotiation act are carried out with a deontic logic engine based on an identity of said at least one of said one or more avatars to which said human speech signals and said non-speech behavior are directed, an identity of said at least one of said one or more negotiation software agents, an action type of said first proposed negotiation act, and a time stamp.
13. The method of claim 5, further comprising, responsive to said disapproving, informing said at least one of said one or more negotiation software agents from which said second proposed negotiation act is obtained why said second proposed negotiation act was disapproved.
14. The method of claim 4, wherein said approving and disapproving comprise comparison to a set of rules.
15. The method of claim 14, wherein said rules enforce turn-taking.
Unknown
September 6, 2022
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.